Quantitative Methods of Predicting Relative Effectiveness of Corrosion

Chapter 17

Downloaded via SWINBURNE UNIV OF TECHNOLOGY on November 27, 2018 at 01:05:04 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Quantitative Methods of Predicting Relative Effectiveness of Corrosion-Inhibiting Coatings on Aircraft Aluminum

K. J. Lewis, J. H. Aklian, and J. D. Zook

Courtaulds Aerospace, International Research Center, Burbank, CA 91504 The aerospace industry visually evaluates results from salt spray and filiform exposures to judge the effectiveness of chromate-inhibited coatings. These methods, however, do not provide enough information for non-chromate-inhibited coatings. Quantitative tests are needed and are described in this paper. Electrochemical impedance spectroscopy (EIS) and galvanic current (I ) are two such tests. EIS response on bonded films measures barrier properties (R ) of coatings. EIS on corresponding nonbonded films measures the electrochemical passivation activity (R ) of the inhibitors they contain. Accurate prediction of field performance requires knowing both R and R . Measuring I with atightlyspacedtitaniumcathode gradually creates the anaerobic acidic conditions of crevice and pit environments. Inhibiting corrosion of aircraft aluminum alloys in these conditions is a rapid predictor of inhibitor performance in real crevice or pit environments. Both methods require comparing non-chromate candidates to uninhibited and chromate-inhibited controls. These tests, along with the traditional ones, will provide the greatest confidence in predicting real world performance. galv

pore

ct

pore

ct

galv

Corrosion-protective paints and sealants for aircraft must protect high strength, coppercontaining aluminum alloys, which are more susceptible to corrosion than pure aluminum. A polymeric coating protects a metal from corroding in two ways. First, it serves as a barrier to water, oxygen, and corrosive ions. Second, it serves as a reservoir of electrochemically active inhibitors with low, butfinite,water solubilities. The following characterize the total corrosion process for a coated metal: 1. Transport properties of a coating or sealant for water, oxygen, and ions (permeability/mechanical integrity) 2. Water collection at the polymer/metal interface (hydrolytic stability or wet adhesion) 3. Availability or escape rates of inhibitors from resin matrix to metal surface (solubility/dispersion/wetting) 4. Activity of inhibitors in passivating a metal surface through adsorption, precipitation, or otherwise plugging pores in a coating 5. Susceptibility of the alloy surface for localized vs. uniform corrosion ©1998 American Chemical Society

Bierwagen; Organic Coatings for Corrosion Control ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

223

224 Number one and two comprise the barrier properties of the coating or sealant. Number five involves the electrochemical activity of the inhibitor(s) toward the active sites on a particular alloy. Three and four contribute to both. An active inhibitor must also not interfere with resin stability or cure chemistry. The aerospace industry uses visually evaluated test methods such as salt spray, filiform, and Prohesion exposures. Fortunately, solvent-borne epoxy coatings have excellent barrier properties, and chromate inhibitors give excellent electrochemical passivation of copper-containing aluminum alloys. Since most current coatings use chromate, only barrier properties differentiate available materials. Environmental regulations mandate removal of solvents and toxic additives, including the carcinogenic chromate inhibitors. Suppliers have been unable to develop ultra high solids materials that deliver film and application performance comparable to current solvent-borne systems. Thus, the aerospace industry is looking to water-borne coatings to solve their VOC problems. Water-borne coatings commonly exhibit weaker barrier properties than solvent-borne coatings. Thus, aerospace must develop new test methodologies to adjust to these technology changes. Further, new methods should enable understanding of the separate contributions barrier properties and the electrochemical activity of the corrosion inhibitors each make. Electrochemical passivation will become more important in water-borne coatings and must be measured accurately. Finally, these methods should extrapolate laboratory data to real world performance without field history. Visual tests for limitedtimesdo not separate the two contributions and may not reliably predict long-term field behavior. Quantitative electrochemical methods must now be added to accepted visual evaluations to improve confidence in predictingfieldperformance from accelerated lab tests. The basic corrosion reactions for aluminum can be expressed as follows: Anode (oxidation, electron loss): Metal E.g., Al° 0

+

-* Metal ΑΓ

+ +

3

e (metal dissolution) 3e

Cathode (reduction, electron gain): 0 + 2 H 0 + 4e_ -* 4 OH (pHt) Or 2 H + 2e -* H t (metal embrittlement) 2

2

+

2

(1) (2) (3)

Water Dissociation: +

H 0 -* H + OH + 3 H 0 -* Al(OH) t + 3 H (C1 ) 2

3

E.g., Α Γ (CI )

+

3

2

3

(pHI)

(4)

Oxygen-rich areas electrically coupled to anaerobic areas provide the cathode reaction supporting continued anode metal dissolution in crevice areas (Fig. 1) (7). The continued formation of metal ions then locally dissociates water to form insoluble hydroxides (Eq. 4). Chloride ions are drawn in to maintain electrical balance while H ions build up. Thus, pH drops, ultimately to an equilibrium level of — 3 for aluminum. The presence of a polymeric film on metal creates a differentially aerated environment under the film as it starts to delaminate. Acidic conditions develop in anaerobic underfilm areas and lead to accelerated corrosion. This occurs in edge filiforms, underfilm blisters, and in pits on bare metal areas exposed upon coating loss. We test inhibitors under acidic conditions and in neutral environments. Others have reported correlation of increased resistance to localized pitting and crevice corrosion in field tests, with measured metal passivation under acidic conditions (2). Disbondment can also occur in alkaline cathode areas, depending on the adhesion mechanism. Small localized pH changes in either direction can come from incipient corrosion in the presence of moisture. These changes can reverse adhesion based only +


225 on polar acid/base interactions (such as from organosilanes). Disbondment allows water to collect at the interface, leading to corrosion. We found strong mechanical adhesion of a coating to aluminum's own high surface area oxide to be very important in preventing corrosion. A good, adhesion-promoting conversion coating achieves this. Others have also documented this need on aluminum, even in the presence of chromate inhibitor in the subsequent primer coating layer (J). Experimental We used 14 methods to compare corrosion inhibitors on aircraft alloys. Testing was done on solutions and on coating or sealant formulation films. These methods include: 1. pH Range immersion exposures of aluminum alloys in inhibited aqueous salt solutions. Visual comparisons are made. 2. Galvanic current measurements (I ) (4) of salt solutions containing dissolved inhibitors. Solutions are held between tightly spaced titanium and bare aluminum alloy electrodes. Quantitative and visual comparisons are made. 3. The method in No. 2, but with inhibitors entering salt solutions from patterned or scribed formulation films bonded to the aluminum. 4,5. Electrochemical impedance spectroscopy (EIS) (5) of dissolved inhibitors in neutral Q T acidic aqueous salt solutions contacting bare aluminum alloys. Quantitative and visual comparisons are made. 6. EIS persistence testing of salt-solution-deposited inhibitors on aluminum alloy. Measurements are made before and after removal of inhibitors from contact solutions. Quantitative and visual comparisons are made. 7-10. EIS testing of formulated films containing inhibitors, (a) where the films are bonded to aluminum alloy, and (b) where they are laid against bare alloy after casting and curing on a nonadhering surface. Neutral or acidic salt exposure solutions are used. Quantitative and visual comparisons are made. 11,12. Salt fog or acidic salt fog exposure (6) of flawed (scribed) formulation films bonded to aluminum alloy. Visual comparisons are made. 13. Prohesion (alternating wet/dry) exposure ( 7) of flawed formulation films bonded to aluminum alloy. Visual comparisons are made. 14. Filiform exposure (8) of flawed coatings bonded on aluminum alloy. Semiquantitative visual comparisons are made, measuring filiform length/frequency. çldv

Inhibitors are screened by dissolving them in salt solution in five of the methods (Nos. 1, 2, 4, 5, and 6). The other nine methods test them in actual formulas. Corroboration among several methods gives confidence that formulas will protect alloys for the life of an aircraft, in a variety of exposure environments. Those environments differ with location on an aircraft and with the aircraft's geographical routes. The above tests also create varying environments, from high humidity to total immersion, and from aerated (neutral/alkaline) to deaerated (acidic) crevice conditions. Inhibitor-Containing Solution Tests. The solution tests are screening tools for finding electrochemically active inhibitors. The pH range immersion test (No. 1) visually evaluates pieces of aluminum alloy sheet immersed in inhibitor-containing aq. NaCl or A1C1 solutions. Inhibitor concentrations are ~ ImM, or saturated for low solubility materials. Solutions are buffered from pH 3-10. A1C1 is used for pH 3, acetic acid/sodium acetate for pH 4-6, nothing for pH 7 and ΝΗ,α/ΝΗ,ΟΗ for pH 8-10. Chromâtes inhibit aluminum alloys for many months, over theftillpH range. Nonchromates, showing protection at low pH for as little as a few days, usually inhibit for 3

3


226

months in neutral conditions. They perform better than uninhibited controls in most other tests as well. The remaining tests with inhibitors added to solution, measure I or are variations of EIS tests. They are described later. flllv

Tests on Formulated Coatings and Sealants. 1,,^ and two types of EIS are run on formulated coatings and sealants to compare and corroborate results of standard salt fog, filiform, and Prohesion exposure tests. Non-chromate inhibitors are always compared to uninhibited and chromate-inhibited controls, as well as to each other. 1^ Measurement. To measure L,^, a titanium cathode is electrically shorted to the active metal, usually Al alloy, in j.5 percent aqueous NaCI solution for up to 1000 hours. Figure 2 illustrates the modified EIS cell used for I measurements. The 1^ method was adapted from one introduced in 1968 for corrosion testing chromatecontaining sealants for aircraft (4). Stainless steel, cadmium plated steel, and carbon composite can also serve as cathodes. Titanium is preferred because of the convenient signal level (/*amp range) and because titaniumrivetsare used extensively on aircraft. ealv

On Solutions and Formulated Coatings and Sealants. In the cell configuration, the two metals are directly shorted electrically, except during the brief periods of measurement. They are separated in the immersion area with "O" ring spacers. Liquid fills the gap through holes in the titanium. When inhibitors are added in the salt solution, bare aluminum is used. With formulated films, the aluminum is coated on the side facing thetitaniumcathode. A pattern of bare area is left to produce the galvanic signal. Uninhibited NaCI solution is used and the inhibitors leach from the film. The "up" side of the uncoated cathode has access to renewing dissolved oxygen. In the closely spaced gap between the metals, oxygen gets depleted, creating a crevice environment. A Keithley picoammeter (Model 485) measures current between anodes and cathodes at programmed intervals. It is connected through a computer-controlled, multichannel scanning/switching device (Keithley Model 706) that controls up to 49 cells. The computer feeds the data to a spreadsheet program and plots current vs. time. EIS Measurement. EIS measurements are made in the cell shown in Figure 2, using the setup with notitaniumcounterelectrode. Instead, a passivated stainless steel counterelectrode and calomel reference electrode are held ~1 inch above the working aluminum electrode with a rubber stopper. A potentiostat (Schlumberger Model 1286), with a frequency response analyzer (Model 1255), makes the measurements. On Solutions. In solution tests, dissolved inhibitors or formula-extracts are added to neutral or pH-adjusted, NaCI solutions. These solutions act on "O"-ring-defined areas of bare metal. Results are judged on the magnitude and stability overtimeof the charge transfer resistance (RJ. R is measured before and after removal of inhibitorcontaining aq. NaCI, followed by replacement with uninhibited aq. NaCI. The retention of high R (> 10 Ω-cm ) after removal of inhibitor-containing salt solution is described as "persistence. " It suggests whether passivated metal will remain passivated after depletion of inhibitor from a coating, i.e., in grooves. Chromâtes form a "persisting," insoluble passivation layer on aircraft aluminum. rt

6

2

rt

On Formulated Coatings and Sealants. EIS is run with two setups to separate measurement of a coating's barrier properties from determination of electrochemical activities of inhibitors contained in the coating. Exposure solutions are 3.5 percent NaCI. To determine barrier properties, EIS is run on fully covering samples of equal


227

Figure 1 - Crevice (or Pit) Corrosion Mechanism (Reproduced with permission from réf. 1. Copyright 1986, The McGraw-Hill Companies.)

Figure 2 - I

galv

and EIS Test Cell Configuration


228 film thickness, bonded to similar substrates. Pore resistance (Rp^J is measured over time. The diameter of the highest frequency semicircle in a Nyquist plot, or the resistance of the lowest horizontal plateau in a Bode plot, yields R ^ . R quantifies the electrochemical activity of inhibitors contained in a formulated film. The exposed metal/electrolyte interface area must be known to determine comparative values. We test cast, cured films. They are simply laid against the substrate and an "O" ring seals around the edges. Once the "nonbonded" film is penetrated, the inhibitor-carrying liquid spreads, by capillary action, over the whole sealed area under the film. Some buckling usually occurs withrigidfilms as they swell. Often some areas do not get wetted, but water marks are then visible, whether corrosion has occurred or not. Dry areas are determined with grid paper laid over the outlines. Simple corrections to "wetted" areas are always less than an order of magnitude. They are at least as accurate as determining pore and delaminated areas in a bonded film, following current literature techniques. These techniques make controversial mathematical assumptions (9, 10, 11) and use curve fitting, or they use complicated optical techniques (72). Errors can be especially large when delaminated areas are orders of magnitude less than the total exposed area of a bonded film. The resistance computed from the diameter of the second highest frequency semicircle in the Nyquist plot for a nonbonded film is adjusted for the correct metal/electrolyte interface area under the film. R for inhibitors leached through the films are then compared accurately. Alternatively, extracts of these cast films can be used to expose bare substrates to determine R . Neither method precisely represents real concentrations because cast films are usually 2-4 times thicker than the films used on aircraft. Nevertheless, the dissolution process from the film will simulate actual use. rt

rt

d

Substrate Effects. Inhibitor corrosion protection must be compared on the same substrates. Large differences in corrosion resistance of the various aluminum alloys (including clad and conversion-coated surfaces), often give more variation in test results for the same coating, than differences between inhibitors (including none), in the same coating, on the same substrate. On the more corrosion-resistant substrates, differences in inhibitor behavior are difficult to discern. Clean Alclad surfaces are pure aluminum and inherently resistant to corrosion, unless contaminated. They differentiate inhibitors poorly. Alloy 2024 pits and crevice corrodes more readily than 7075. Both are active because of their copper sites. As bare substrates, both differentiate inhibitors well. Matrix Effects. Inhibitors should always be compared in the same resin matrix, loaded to the same total pigment volume concentration (PVC). This holds water permeability and adhesion of the polymer phase as constant as possible. The kinetics of inhibitor escape from a resin into the aqueous environment is an important parameter, especially in certain use environments. Damage-exposed bare metal areas, next to coated areas, should be protected as well as undeifilm areas. Healing of established pits and crevice areas is difficult, even for chromate. Fast-leaching materials generally protect better, by passivating bare areas before corrosion processes become well established. Results and Discussion Igaiv îgaiv test cell configuration accelerates the development of acidic crevice conditions. This happens after ~100-1000 hours in the narrow, air-deficient gap between the 0"-ring-spaced anode and cathode metals. Figure 3 shows an I data comparison for noncuring sealants. The effectiveness of inhibitors in suppressing galvanic current is compared, both initially and with up to 1000 hours of exposure. T

h

e

H

galv


229 Acidic conditions gradually develop in prolonged exposure, unless corrosion is totally suppressed. Acidity accelerates the corrosion rate. Four of the samples contained (1) no inhibitor, (2) a low level of SrCr0 , (3) a neutral-inhibiting, non-chromate system, and (4) an acid-condition-inhibiting, nonchromate system. These four were in the same noncuring sealant matrix. The fifth sample contained BaCr0 at a higher level. A slower inhibitor dissolution rate is seen in the expanded X-axis plot on therightof Figure 3. 4

4

Low pH Current Acceleration. The measured pH of gap liquids is always 3-4 in uninhibited or poorly inhibited I cells opened after the current rate has steeply climbed. The aerated side of thetitaniumcathode provides an accelerated cathode reaction to drive development of anodic crevice conditions in the confined gap. The accelerating currents of such samples come from both building acidity and increasing exposure area from edge spots of underfilm corrosion. The latter occur especially on 2024 alloy. Hydrogen reduction ultimately occurs. Accelerated currents then level off or fall if the gap becomes gas-bound or clogged with corrosion product. Chromate-inhibited controls rarely climb in current unless too low a level is used. The pH usually stays steady at the starting level of 6-7. However, crevice blisters have been seen with inadequately-inhibited chromate systems, following inhibitor depletion. galv

Acid-resistant Inhibitors. Some non-chromate corrosion inhibitors resist I cell crevice conditions longer than others. We found such results to follow thetimingfor appearance of acidic crevice blisters in salt spray exposure. We compared the I results of Figure 3 with salt spray results on the same noncuring sealants, on 2024 alloy. Acidic crevice blisters formed on the uninhibited sample by 500 hours. They appeared in the neutral-condition-resistant, non-chromate sample by 1200 hours. They did not appear with the acid-resistant non-chromate system until 2000 hours. At higher loading, in cured sealant, a related, acid-resisting, non-chromate system lasted over 4000 hours in salt spray. No crevice blisters formed. This salt spray sample also exhibited no corrosion of adjacent bare metal areas within 1/8 inch of the sealant edge. I results for this material are shown in Figure 4. EIS measurements of a nonbonded film of it are shown in Figure 5. Both tests indicate good electrochemical passivation by the inhibitor system. Figure 6 shows I data for three other, related, acid-resisting, non-chromate inhibitors in an amine-cured, water-borne epoxy primer. Chromate-inhibited and uninhibited controls are included. Long-term crevice current was suppressed. galv

gftlv

galv

galv

Inhibitor Passivation of Bare, Damaged Areas Next to Coating or Sealant. Surface Transport Properties of Inhibitors in a Damp Environment. SrCr0 and 4

CaCr0 , in the uncured and cured sealants in Figures 3 and 4, prevented blisters up to 2000 and 4000 hours, respectively, in salt spray. Pits formed in l/4"-wide bare metal areas next to the sealants, however, as early as one week on alloy 2024. Alloy 2024 is prone to localized corrosion. *CmOmate does not migrate well onto bare areas unless the matrix is immersed. The acid-resistant non-chromate system in Figures 3 and 4 actually protected adjacent bare metal areas against pitting better than chromate in these sealants. They exhibited surface migration in a damp, but not immersed environment. Protecting adjacent, exposed damage sites is an issue to consider since such sites will pit and corrode vertically into the metal, compromising its structural integrity. Daimler Bena Aerospace Airbus in Europe does a crevice test where both vertical and lateral corrosion penetration are measured from a damage site. Inhibitors able to migrate from a coating or sealant to passivate adjacent bare areas in a "damp but not immersed" environment provide a performance advantage. 4


230

Exposure Time in 3.5% aq NaCi (hours) 1- Uninhibited

2- Lo level SrCr0

4

3- Non-chromate (neut resist)

4- Non-chromate (acid resist) Figure 3 - I

galv

5- Hi level BaCrQ

4

Measurements, Noncuring Sealant, 2024 Alloy

3E-05

6Ε-05

r

"7—

5Ε-05 CO

c S 2E-05 H +-» c ω h.

4Ε-05 i l l

;

ν /

ι

te

ι

3Ε-05

ϋ

2Q24ALLOY

S > 1E-05 4/

2Ε-05 -1

7075 ALLOY

(Ο

ο

1Ε-05

0

0 200 400 600 800 0 200 400 Exposure Time in 3.5% aq NaCI (hours)

1- Uninhibited Figure 4 - I

2- C a C r 0

galv

4

600 800

3- Non-chromate (acid resist)

Measurements, Cured Sealant, Both Alloys


231

Figure 5 - EIS Measurements, Nonbonded Polythioether Sealant, 7075 Alloy


232 EIS. On Solutions. EIS testing of inhibitor solutions at both neutral and acidic pH was used as a screening tool on bare metal substrates. Along with pH range immersion tests and I measurements on solutions, solution EIS identified electrochemically active inhibitors and combinations for further testing in formulas. Beyond looking for high initial (> 10 Q*cm ), we screen for "persistence" of high R . Recall we remove the inhibitor-containing solution after exposurefrom1 day to 1 month, replace it with uninhibited NaCI, and remeasure R* over time. Figure 7 shows EIS data for SrCr0 and for both persisting and nonpersisting nonchromate inhibitors, in NaCI solution, on 2024 Al alloy. For chromate and for the persisting non-chromate system, initial R is > ΙΟ Ω-cm , and it remains essentially unchanged after removing the inhibitor(s). With a nonpersisting inhibitor, a high initial R can last for weeks in an undisturbed, sufficiently concentrated solution. The equilibrium adsorption is destroyed, however, when the inhibitor-containing NaCI solution is removed and replaced with just NaCI. R drops to < 10 Ω-cm , and the aluminum then corrodes as if no inhibitor were present. g?tv

6

2

ct

4

6

2

rt

ct

5

2

rt

On Coatings and Sealants. Recall inhibitors should be compared in the same resin matrix, at the same PVC (or PVC/CPVC ratio), at equal thickness, on the same alloy substrate. Meaningful conclusions can then be drawn as to the inhibitor's contributions to both barrier properties and electrochemical passivation activity. Measured resistance is inversely proportional to corrosion current, so both R ^ and R should be as high as possible. Generally, the lower the resistance to pore formation (low Rpore), the poorer a coating system performs, especially if the measurement is on a primer/top coat combination. However, a high R* within the pores of a primer (at the metal/electrolyte interface) can go far in overcoming a pore formation problem with a "water-sensitive" inhibitor. Good corrosion protection can still result, provided enough inhibitor is present and good adhesion is maintained, especially if that adhesion is mechanical. We found this to be the case with all the SrCr0 -loaded epoxy primers we tested. Not surprisingly, we also found the loading level, dispersion quality, and water solubility of added pigments to greatly influence R ^ of the primers we tested. Figure 8 compares uninhibited and SrCr0 -inhibited, bonded primers, made from the same solvent-borne, amine-cured epoxy formula, at the same total PVC. Both were coated on 2024 alloy at the same — 4-mil thickness, without topcoat. The barrier performance of the SrCr0 -loaded sample (determined from R^J was seen to degrade more rapidly than that of the uninhibited one. In the uninhibited sample, the pigment volume which was occupied by SrCr0 in the original formula, was replaced with less-water-sensitive, hydrophobic silica. This improved R compared to the original SrCr0 formula. At the sametime,R*, represented by the second semicircle diameter in each of the Nyquist plots, is uncorrected for actual metal/electrolyte exposure area since it is unknown. The exposure areas for the chromate sample have to be larger because of their lower R ^ values (taken from thefirstsemicircle diameters). Thus the real, areacorrected Rct values for the SrCr0 sample would be larger multiples of the measured resistances than those for the uninhibited sample. Since the measured R values for the chromate sample are already much larger without correction than for the uninhibited one, the real differences are huge. rt

4

4

4

4

pore

4

4

rt

On Topcoats (or Unicoats). EIS can be used to measure barrier properties (porosity) and water uptake of topcoats. Porosity (R^J changes will profile the UV resistance of an exterior coating. EIS will corroborate measured gloss changes on films exposed to normal or accelerated weathering.


233

1 - SrCrCV original, after ldy &after lmo w/NaCl only Frequency (Hz) 2 - Persisting non-chrome, original io~ io~ icr 10° ί ο ί ο i o 3 - Persisting non-chrome, after 4hr w/NaCl only 7 4 - Persisting non-chrome, after lmo w/NaCl only 5 - Non-persisting non-chrome, original 6 - Non-persisting non-chrome, after 4hr w/NaCl only 3

2

1

1

2

3

Log

1000000

I ZI

1

10~ IO" I O i o 1 0 1 0 w (radians/sec) 2

10 -90

1

0

2

to

3

4

Frequency (Hz) 10~ 10" 10° 1 0 1 0 1 0

-3

2

1

1

2

3

Angle (deg) 1000000

2

1

1

2

3

10" IO" 10° I O 1 0 1 0 1 0 w (radians/sec)

Figure 7 - EIS Persistence Measurements - SrCr0 and Persisting and Nonpersisting Non-chromate Inhibitors, in Solution on 2024 Alloy 4


4

234

Frequency (Hz) 0

1

2

3

4

îcrtcr^crto ίο ίο ίο ίο ίο

w

(radians/sec)

Figure 8 - EIS Measurements - Bonded Epoxy Primers, 2024 Alloy


5

235

For topcoats providing corrosion protection in addition to cosmetics, the issue of porosity changes in terms of barrier properties is critical. Another paper in this symposium (75) presented data comparing conventional salt fog, filiform, and Prohesion results with corrosion data from outdoor exposure at various USA sites. There was no correlation at all (coefficients were 10 Ω · cm , suggesting good passivation. Recall this sealant performed well as a bonded film in salt spray. s

2


236 Table I shows bonded and nonbonded film data on uninhibited, SrCr0 -inhibited, and non-chromate-inhibited versions of an amine-cured, water-borne epoxy. They were air-dried and measured without topcoats. While film thicknesses were not exactly the same, all were penetrated quickly. A low was measured on each bonded coating, from the high-frequency-semicircle diameters of Nyquist plots, or the high-frequency, lowest-plateau values of Bode plots. An area-corrected was determined for each nonbonded coating from the Nyquist next-lower-frequency-semicircle diameters. Values for both chromate- and nonchromate-inhibited films were 1-2 orders of magnitude higher than for uninhibited films, even after 50 days. Where values were > 10 , it was difficult to extrapolate the semicircle diameters accurately. Measuring to very low frequency to improve accuracy was often not possible. Noise from "autointegration failure" in the software computations begins to compromise the low frequency data when R is very high. 4

7

rt

Conclusions. The aerospace industry currently uses visually evaluated salt spray and filiform exposures to judge the effectiveness of chromate-inhibited coatings. Current coatings differ mainly in their barrier properties for protecting aluminum alloys since most contain SrCr0 , although some BaCr0 and ZnCr0 are used. Salt spray and filiform alone do not provide enough information for non-chromates. Quantitative tests such as electrochemical impedance spectroscopy (EIS) and galvanic current measurement (I ) are needed. To make valid rankings of non-chromate inhibitors, they should be compared against uninhibited and chromate-inhibited controls, as well as to each other. Matched matrices, with equal pigment volume concentrations, should be used at the same thickness, on the same selection of aluminum alloy substrates. A dichotomy exists in formulatingfilmsfor corrosion protection of metal. Loading a hydrophobic polymer film with hydrophobicfillerscan preserve its water resistance and barrier properties. Conversely, loading a hydrophobic polymer with effective inhibitors that are water-soluble enough to enter the electrochemical environment to stop corrosion will always compromise barrier properties to some extent. Barrier properties are thefirstline of protection against corrosion and should be at a maximum as long as possible. EIS response on bonded films measures barrier properties (RpoJ of inhibited coatings. EIS on corresponding nonbonded films measures the electrochemical passivation activity ( R J of the inhibitors themselves. Accurate prediction of field performance requires knowing both R ^ and R . Both should be as high as possible. Rpo,, for a non-chromate coating should be no worse than for a similarly-loaded chromate control. However, it may be poorer than for an uninhibited control, as is SrCr0 . R^ for a non-chromated coating should exceed R* for an uninhibited control, and should approach a similarly-loaded chromate control. When using filiform testing for corroborating electrochemical tests, relative humidity should be kept high enough for some liquid water to form during cycling. This appears necessary for electrochemical properties to be a factor in the corrosion resistance measured in this test. Lower humidity favors barrier property domination, especially while HC1 vapor is present in a film. Inhibitor passivation of aircraft aluminum alloys in acidic solution is useful in predicting suppression of localized corrosion. Measuring I ^ between aluminum alloy and atightlyspacedtitaniumcathode is recommended for 100-1000 hours, to accelerate development of acidic crevice conditions. With this method, inhibited formulas can be ranked as to capability to retard crevice and pit corrosion. In salt spray, inhibitors that migrate from a formulation and passivate adjacent exposed metal areas, in a damp environment, can provide a performance advantage. 4

4

4

galv

rt

4


237 Table I - EIS Coatin g Measurement Results, Water-borne Epoxy Primers •cm ) Film Thickness (mils) R,™ (Ω-cm ) from Bonded from Nonbonded Inhibitor Exposure Bonded 1 Nonbonded Films Films Time (24 cm ) (1 cm ) (24 cm ) 1x10 1 hr None 1.5-1.8 5-7 1.5x10 >10 3 hr 2x10 7x10 1 day 3.5x10 2 day 7x10 4x10 2500 6 day 3x10 2x10 1000 50 day 3x10 1 hr 1.8-2.0 8-9 SrCr0 lxlO >10 3 hr 2X10 5x10* 1 day 4X10 2 day >10 3-4xl0 1000 6 day lxlO 2xl0 500 50 day 8x10* 2.5 6-10 1 hr CA l-2xl0 5x10* 3hr Non>10 8000 chromate 1 day lxlO 2 day lxlO lxlO 1500 6 day lxlO 8x10 1200 50 day 2

2

1

2

2

2

s

s

s

s

s

s

s

s

s

s

s

4

s

6

6

6

7

8

8

7

7

7

7

7

7

s

7

Acknowledgments. The authors thank T. Dang at Courtaulds Aerospace for his fine artwork in Figures 1 and 2. We also thank Courtaulds Coatings, UK for providing us with EIS equipment for this project. Literature Cited. 1. Fontana, M.G., Corrosion Engineering, McGraw-Hill, 54 (1986). 2. Szklarska-Smialowska, S., Reviews on Corrosion Inhibitor Science and Technology,NACE,I-9-1(1989). 3. Evans, A.T., Scantlebury, J.D. and Callow, L . M . , Proceedings of the Symposium on Advances in Corrosion Protection by Organic Coatings II, Electrochem. Soc., 95-13, 266 (1995). 4. Morris, L . , "Corrosion of Galvanic Couples Found on Aircraft and the Effects of Inhibitive Sealants", Report to NACE Technical Unit Committee T-9c, Fuel Systems Corrosion, Cleveland, OH (1968). 5. MacDonald, J.R., Ed., Impedance Spectroscopy, Wiley & Sons, 1987. 6. ASTM B 117-85. 7. Cremer, N.D., "Prohesion Compared to Salt Spray and Outdoors", Presentation to Federation of Societiesfor Coatings Technology, 1989 Paint Show, UK, Reprint from Q-Panel Co., 1989. 8. ASTM D 2803-70 (reapproved 1974) and MIL-P-23377G (1994). 9. Haruyama, S., Asari, M . and Tsuru, T., Proceedings of the Symposium on Corrosion Protection by Organic Coatings, Electrochem.Soc.,87-2,197 (1987). 10. Mansfeld, F. and Tsai, S., Corrosion, 47 (12), 958 (1991). 11. Kendig, M . , Mansfeld, F. and Tsai, S., Corrosion, 47 (12), 964 (1991). 12. Kendig, M . , Ryang, H . , Liao, T., Cunningham, M . and Jeanjaquet, S., "Corrosion Induced Adhesion Loss-Low Volatile Organic Content (VOC) Coatings", Final Report to Material Interfaces Div., Office of Naval Research, Contract#N00014-90-C-1011,March 1993. 13. Simpson, C.H., "The Replacement of Chromate Inhibitors in Protective Coatings", this volume (1997). Bierwagen; Organic Coatings for Corrosion Control ACS Symposium Series; American Chemical Society: Washington, DC, 1998.

Quantitative Methods of Predicting Relative Effectiveness of Corrosion

Recommend Documents